Reading the Curry Mile – a Case Study

! Report! 2014!

! The$contents$of$this$report$are$the$intellectual$property$of$the$ authors.$No$part$of$this$report$may$be$circulated$or$reproduced$ without$ explicit$ permission$ from$ the$ authors,$ or$ from$ the$$ School$ of$ Arts,$ Languages$ and$ Cultures$ at$ the$ University$ of$$ Manchester,$ Oxford$ Road,$ Manchester$ M13$ 9PL,$ United$ KingI dom.$

Reading!the!Curry!Mile!–! Language!Use!in!the!Linguistic Landscape!of!Rusholme,!Manchester

Leonie$Elisa$Gaiser$ 2014$ $ $ $ $ $ $ $ $ A$dissertation$originally$submitted$to$$ the$University$of$Augsburg,$Germany$ $for$the$degree$of$Bachelor$of$Arts$$ $

Table of Contents

1$ Introduction ...... 1$ 2$ Approaching Linguistic Landscape ...... 3$ 2.1$ Towards a Definition of Linguistic Landscape ...... 3$ 2.2$ Authorship in the LL ...... 4$ 2.3$ Unit of Analysis ...... 5$ 2.4$ Functions of Language in the LL – A Critical Perspective ...... 6$ 2.4.1$ Landry and Bourhis – Informational versus Symbolic Functions ...... 6$ 2.4.2$ The Commodification of Language ...... 7$ 3$ Functions of Language in the LL – A Linguistic Foundation ...... 8$ 3.1$ LL as a Form of Communication ...... 9$ 3.2$ Functions of Language – A Cognitive-Sociolinguistic Perspective ...... 9$ 4$ Reading the Curry Mile – A Case Study ...... 11$ 4.1$ Research Objective ...... 11$ 4.2$ Methodological Considerations ...... 12$ 4.2.1$ Survey Items ...... 12$ 4.2.2$ Internal Variables of Signs ...... 15$ 4.3 Research Area and Fieldwork – The Curry Mile ...... 18$ 5$ Language Use on Signs in the Curry Mile – Quantitative Findings ...... 20$ 5.1$ Top-Down Items ...... 21$ 5.2$ Transgressive Items ...... 22$ 5.3$ Bottom-Up Items ...... 22$ 5.3.1$ Language Use across Domains ...... 25$ 5.3.2$ Language Use Outside and Inside Establishments ...... 25$ 5.4$ An Initial interpretation ...... 26$ 6$ Language Functions in the LL – A Qualitative Perspective ...... 29$ 6.1$ Language in the LL - An Expression of Identity and Authenticity ...... 29$ 6.2$ Commodification of Language and Detachment of Ethnicity ...... 31$ 6.3$ Monolingualism on External Signage ...... 33$ 6.4$ Minority Languages in the LL – A Means of Communication ...... 35$ 6.5$ Language Functions and Reader-Dependency ...... 36$ 6.6$ The Functions of English in an Anglophone Context ...... 39$ 6.7$ Language-Contact and its Functions ...... 41$ 7$ Discussion ...... 43$ 7.1$ The LL and Ethnolinguistic Vitality ...... 43$ 7.2$ Language Use and Functions across the LL ...... 43$ 7.3 The Functions of Languages as Gradient Features ...... 45$

8$ Conclusion ...... 48$ 9$ Bibliographical References ...... 51$ 10$ Appendix ...... 55$

1 Introduction

Walking through the city, one is confronted with a variety of languages in multiple forms and functions. Since it is visually omnipresent on commercial signs, leaf- lets, T-shirts or advertising on vans and buses, written language considerably marks the public space. While ordinary passers-by are usually not fully aware of it, scholars have in the last decades devoted increasing attention to language use on signs in urban areas. Understanding such linguistic landscapes as “complex indexes of source, addressee, and community” (Collins/Slembrouck 2007: 335), it becomes evident that their investigation is extremely useful in various fields and for distinct purposes. Originally, the term linguistic landscape was used to refer to the more general linguistic situation of a particular area, describing the presence and use of languages in a given region, city or country. Kress and Van Leeuwen (2006) turn their attention to the term itself and note its revealing etymology, illustrating the dynamic character of the linguistic landscape (henceforth LL). “While to the casu- al beholder a landscape simply is, and may even have a timeless appearance […], it is in fact a product of social action and of social history, of human work on the land, on nature: -scape, with its relatives to shape in English and schaffen (both ‘to work’ and ‘to create’) in German, indicates this” (Kress/Van Leeuwen 2006: 33). Similarly, new texts may be added to the LL, while other signs may be removed; hence, it is a phenomenon that is subject to constant re-shaping by a wide range of legal, social, economic, cultural and even emotional factors. Particularly in areas where a variety of ethnic groups interact, publicly displayed language and script can be understood as a negotiation of identity. Thus, the LL seems to be much more than written communication of factual information. Moreover, “foreign language use is far from being a kind of emotionally ‘neutral’ facet in language contact settings” (Haarman 1989: 53). Previous research has indeed acknowledged the symbolic values of language on signs, but many approaches adopt a rather limited perspective. The varied motivations and incentives possibly influencing the sign-producers’ linguistic choices have not yet been adequately investigated, neither have those meanings that may possibly be ascribed by the reader been explored. In a multilingual environment, several languages play distinct roles in social interaction, and the mere act of choosing language and script choice is to be regarded as highly indexical in itself. Likewise, the recipient is an essential participant in LL communication and deserves particular attention. Hence, a participant- 1 oriented and contextualised approach investigating language use on individual signs seems necessary to determine the functions of language in the LL. This paper presents a case study into the functional aspects of language choice in the linguistic landscape of the Curry Mile in Manchester, UK. Basing interpretations on a quantitative analysis of language use on signs in the study area, it will be investigated whether patterns of language choice can be linked to the types of sign-producers and domains. It is the aim of his paper to reveal the various dynamic functions of language; it will be analysed to what extent certain languages for certain viewers function as a means of communication, while others, in the given context, seem to be used mainly for symbolic functions. A characteristic feature of the present study is the analysis of written language use inside local establishments in addition to written language that can be seen from the streets. The subsequent chapter is concerned with the general concept of linguistic landscape. It provides an overview of the existing literature and presents some of the parameters and methodologies most relevant to this study. Providing the sci- entific basis of the present research, it will then be dealt with the abovementioned linguistic theories in relation to the LL. After this, a definition of the research questions resulting from the objective of this paper will be presented, and the established methodology is explained before the paper turns to the findings of the study. The first goal of the paper is to develop an approach that can do justice to the complex interplay of factors influencing the functions of language in the LL, an investigation of which is the main purpose of the present research. It will first be analyzed which languages and scripts are present in the LL of the Curry Mile, and to what extent this reflects the linguistic situation of the area. Quantitative analyses of the collected data relating language use to domains and the signs’ direct context provide a basis for the micro-level analysis of the most representative signs. In the qualitative investigation, special attention is directed to the sample of commercial bottom-up signs. The final aim is to explore whether for certain readers, certain languages play particular roles in the LL of the Curry Mile. On the basis of these results, answers to the research questions are formulated and conclusions are drawn. I will claim in this paper that languages cannot be linked to particular sets of values, neither are they limited to certain functions. Signs do not simply create meaning in isolation; in fact, readers interpret texts and relate this to sign-writers, the signs’ placement and to how they work together with other items in the LL.

2 Approaching Linguistic Landscape

The past decade has seen a rapid development of LL research, and researchers have become increasingly aware of the fact that counting the languages and analysing combinations on signs will not provide much insight into their functions in the public sphere. The following chapter will elaborate on the concept of linguistic landscape, laying out the foundation for the present approach. Moreover, it will present an overview of the most significant studies conducted in the field, and it will discuss the central methodological considerations that are relevant also for this paper. Finally, previous approaches to the functions of language in the public sphere will be outlined and critically examined.

2.1 Towards a Definition of Linguistic Landscape

Although particularly the first LL researches are based on quantitative and rather limited analyses, some of the early studies deserve mention.1 The paper of Ros- enbaum et al. (1977) presents the first study whose structure and character re- sembles the present concept of LL research. The researchers determine the prominence of Roman script on signs in a busy street in Jerusalem. Signs were counted, categorized according to the types of products and classified on the basis of the scripts present on the individual signs. Additionally, Rosenbaum et al. relate their findings to results gained from conversations on the streets and interviews. Another major study in the research field is done by Spolsky and Cooper (1991), who investigate publicly visible forms of written language and the motivations behind code choice. The first use of the term linguistic landscape in the sense used in the present paper is commonly attributed to a seminal paper written by the Canadian researchers Rodrigue Landry and Richard Bourhis (1997). Their pioneer work helped LL research to gain authority as an autonomous discipline, and it provided also the arguably most quoted definition of linguistic landscape: The language of public road signs, advertising billboards, street names, place names, commercial shop signs, and public signs on government buildings combines to form the linguistic landscape of a given territory, region, or urban agglomeration. (Landry/Bourhis 1997: 25)

Their approach still serves as theoretical grounding for a number of studies, but the concept the term linguistic landscape refers to varies in scope from researcher to researcher.

1 Cf. Backhaus (2007) for a comprehensive overview of previous approaches to the LL. 3

Shohamy and Waksman (2009) present a view that goes far beyond the original understanding, encompassing discourses that potentially emerge in public spaces. Apart from the texts, images and objects themselves, their placement in time and space and human beings are included. Other researchers expand their research to questionnaires, interviews and participant observations, or become even personally involved in the community (Malinowski 2009). Reh (2004) introduces the parameter of spatial mobility, focusing on the extent to which LL items are physically fixed. Accordingly, signs may be “stationary, as in the case of buildings or large signboards, or movable, as in the case of newspapers, T-shirts, or books, and transitional forms also exist” (2004: 3). Apart from controversies surrounding the objects to be analysed, the term linguistic landscape itself has been widely discussed, and a number of alternative coinages have risen. Gorter (2006) introduces the term multilingual cityscape, emphasising that the topic of interest is usually written language use in urban areas. His term, however, excludes the possibility of monolingualism on signage. Jaworski and Thurlow (2010) maintain that a differentiation between linguistic landscapes and semiotic landscapes is necessary. The authors argue that “visual images, nonverbal communication, architecture and the built environment” (Ja- worksi/Thurlow 2010: 2) interact with written language. Indeed, the terms linguistic cityscape and semiotic landscape seem to be more appropriate and exact, and the latter represents an argument central also to this paper. However, it was decided to apply the term linguistic landscape in the context of this study as it is the most widespread and established term. The analysis of the LL and an adequate codification of visible signs presents a number of difficulties, which accounts for the variety of parameters and methodologies that have been defined. The two most significant concepts that are relevant to the present case study will be illustrated in the following.

2.2 Authorship in the LL

A notion that has been addressed in most LL studies is the question of authorship. One of the basic differentiations is that between governmental and private signage, distinguishing between a public and a private source or originator of LL items. Various terms designating these categories have been introduced, with the most frequent being official and non-official (Backhaus 2006), governmental and non-governmental (Huebner 2006), or top-down and bottom-up (Ben-Rafael et al. 2006) signs. The former category of each of the conceptual pairs refers to signs that are under direct control of national or local policies. They belong to governmental, 4 religious, educational, cultural, medical, social or legal institutions; examples include public municipal signs, traffic signs and displays of street names. The latter term refers to the group of non-official signs and typically includes “signs set up by autonomous actors such as shop owners, companies or other private enter- prises” (Backhaus 2006: 27). Compared to top-down signs, the state may exert significantly less control over language choices of bottom-down signs. Several researchers criticise the twofold distinction between top-down and bottom-up signage (cf. Coupland 2010; Huebner 2006). Huebner draws attention to problems occurring with regard to store chains, which may impose their own language policy and determine linguistic choices for their signage (cf. 2006: 74). In other studies, more fine-grained approaches have been suggested. Scollon and Wong Scollon (2003) differentiate between four general discourses commonly found in the public sphere of urban areas: “official regulatory discourse”, “infra- structural discourses”, “commercial discourses” and “transgressive discourses” (cf. Scollon/Wong Scollon 2003: 181). The last category includes stickers, printed and handwritten notes, graffiti or other forms of verbal text written on or attached to walls or posters without authorisation, and it is adopted also in the framework for the present study.

2.3 Unit of Analysis

Another methodological problem frequently arising in LL research is the question of what constitutes the unit of analysis. This is indeed a complex task, and a generally acknowledged definition has not yet been established. Some researchers count signs individually, regarding each sign, poster or sticker as unit of analysis. Consequently, a business has several signs if there are distinguishable items displaying language on windows and the façade. An advantage of this approach is that the signs are clearly countable, which facilitates a comparison between studies. However, it can be assumed that readers view the signage of each shop or establishment as an entity and regard it as the product of a single producer. A similar argument is brought forward by Cenoz and Gorter, claiming “that all the signs in one establishment, even if they are in different languages, have been the result of the languages used by the same company” (2006: 71). Accordingly, each commercial LL sign is seen in its context of the entire shop signage it belongs to, and it is thus regarded as part of a “larger whole instead of being clearly separate” (Cenoz/Gorter 2006: 71).

As mentioned above, the present research devotes particular attention to the functions of language on signs in the public sphere. Previous approaches with similar focus on language functions in the LL will be critically examined in the following.

2.4 Functions of Language in the LL – A Critical Perspective

The functions of language, as well as the function of the multilingual LL as an entirety, have been of particular interest in several LL studies; however, interpretations and investigation has been limited so far. Many fail to do justice to the varied and user-dependent factors influencing language choice on the one hand, and the functions of language on the other hand.

2.4.1 Landry and Bourhis – Informational versus Symbolic Functions For Landry and Bourhis (1997), whose work has influenced a great number of researchers in the field, language choice in the LL has two main functions: an informational and a symbolic function. Elaborating Spolsky and Cooper’s (1991) view, it is argued that the LL “serves as a distinctive marker of the geographical territory inhabited by a given language community, […] inform[ing] in-group and out-group members of the linguistic characteristics, territorial limits, and language boundaries of the region where they have entered” (Landry/Bourhis 1997: 25). Accordingly, code choices visible on LL signs serve an informational function inasmuch as they indicate which languages are used for communication in the investigated territory. Landry and Bourhis maintain that the LL can be seen as a concrete reflection of the sociolinguistic composition of the speech communities inhabiting the area (cf. 1997: 26). Apart from this, the presence or absence of one’s language in the LL “can serve a symbolic function that is affectively charged and that can complete the information function” (1997: 27). The inclusion of the in-group language on signs can contribute to the positive social identity of communities. This is related to one of the central points of Landry and Bourhis’ work, which is to connect the notion of LL with the concept of ethnolinguistic vitality. In their view, language use on official and non-official signs indexes the vitality, strength and control of that language group on institutional and demographic dimensions. I argue in this paper, however, that this understanding of the LL as a marker of ethnolinguistic vitality is questionable. First, it is problematic to claim that the LL serves to delineate linguistic boundaries between local communities. Tourists, visitors and professionals who frequent the area may have a considerable impact

6 on the shaping of the LL, and their knowledge of languages might differ from the residents’. Second, language is not necessarily used as a means of communication, but it may serve other purposes such as the expression of values, beliefs and identity (cf. below). Hence, the presence of languages in a multilingual LL does not imply that these are actually used by the speech community in social interaction. Landry and Bourhis’ view of the LL has recently been challenged by other researchers who note that “the degree of prominence of a language in a particular site is not necessarily the most accurate indicator of the ethnolinguistic vitality of its speakers” (Jaworski/Thurlow 2010:10 f.; cf. Kallen 2010; Coupland 2012). Moreover, it has also been pointed out that the “Landry-Bourhis approach sees LL as ‘given’ context of sociolinguistic processes and thus does not focus on the very factors which give shape to LL with limited consideration, if any, to the dy- namics of the LL” (Ben-Rafael et al. 2006: 8). While Landry and Bourhis focus on the informational and symbolic values of language use in the entire LL, other researchers turn their attention to the functions of languages on commercial signs. Functional aspects of language on commercial signs will be elaborated on in the following part of this paper.

2.4.2 The Commodification of Language

Focusing on language use on individual signs, researchers have similarly pro- posed a dichotomy between the informational and the symbolic function; it is emphasised that the main purpose of language is not necessarily to communicate information. Accordingly, particular languages are in some contexts not used to transmit factual information, but chosen mainly for their connotational values. It is argued that certain linguistic repertoires are associated with particular cultures and groups of people, and they asare stereotypically connected to respective competencies, products and social practices (cf. Leeman/Modan 2009). Kelly-Holmes pays particular attention to the symbolic functions of language and claims that frequently, „languages are used for effect or form rather than information or content” (2005: 67). In this context, the author suggests the concept of the country-of-origin effect. Accordingly, languages are chosen for their associations with particular cultures that are believed to have expertise in the relevant product or service area. Cheshire and Moser (1994) investigate the presence of English in non- Anglophone countries and maintain that the language “sometimes functions as a cultural symbol – in other words, not as a system of signs, but as a sign in itself” 7

(1994: 451). Similarly, Haarmann (1989) argues that major foreign languages are associated with specific ethnocultural stereotypes, neglecting their ideational content. Accordingly, a definable range of meanings cluster around languages, and their function as a symbol subordinates its referential content to other functions (cf. Haarmann 1989: 10 ff.). It is maintained that a language is used mainly for its ideational meanings or its symbolic value. Indeed, language is not always intended to communicate referential content. Regarding language on signs mostly as symbolic design element or ornament rather than as a means of communication, Landry and Bourhis’ initial understanding of the LL as a direct reflection of the ethno-linguistic composition of a given area should hence be questioned. Turning to the approaches presented in this section, it is in the present paper not disputed that for a text on a sign, certain language functions may be given greater emphasis than others; yet, caution must be taken when differentiating ideational and symbolic functions of particular codes in the LL. It seems that the twofold distinction cannot capture the variety of functions language may fulfil; more importantly, it is questionable whether languages can be assigned either informational or symbolic power. Most of the previous studies on written language in the public sphere have been far too inaccurate and imprecise. Linguistic foundation is often missing, and several aspects have not been taken into account (cf. Kelly-Holmes 2005, Cheshire/Moser 1994; Landry/Bourhis 1997). With the purpose of comprehensively analysing language-in-use and answering the questions dealt with below, it seems necessary to relate relevant, traditional linguistic theories to language in the LL.

3 Functions of Language in the LL – A Linguistic Foundation

The present LL analysis draws on well-established functional approaches and more recent theories to language in use. These combine to a solid theoretical basis for interpretation. Before presenting the quantitative results and insights gained, it seems necessary to introduce some fundamental concepts underlying this study. In combination, these will provide a basis for the methodological framework applied to the collected data.

3.1 LL as a Form of Communication

In the present paper, the LL is conceived as a “site for the negotiation of meaning between the potential intentionality of the creator and the potential readings of the target audience” (Agnihotri/McCormick 2010: 55). The use of written language in the public sphere is understood as a form of communication between sign-producers and the audience, who “[…] are jointly engaged in creating meaning and handling social relations” (Bublitz 2012: 153).2 Indeed, readers of a signs do most frequently not have direct access to the sign-producer, and meaning cannot literally be negotiated with the writer of the message. Yet, meaning-making can be understood in a metaphorical sense: it “takes place on the plane of cognitive interactivity” (cf. Bublitz 2012: 160) as readers make assumptions about what could be meant. Since the interpretation of the LL depends on individual recipients, it seems impossible to attribute certain languages particular sets of functions and values. Furthermore, I maintain in the present paper that LL signs must be analysed in relation to their direct and indirect environment. “[A]ll reading [of signs] is a contextualized interpretative practice that draws on diverse frames of interpretation” (Collins/Slembrouck 2007: 349); hence, it is necessary in LL research to consider contextual factors that possibly influence interpretation. Similar to the reciprocal character of meaning-negotiation, the construction of context is in this paper understood as a dynamic process. Context itself is viewed as a highly flexible phenomenon that arises in communication. It is not given beforehand, but participants themselves decide which of the provided recourses they relate to language and turn into context.3 Having outlined the hermeneutic, collaborative and contextual theory underlying this paper, particular attention will now be paid to the varied functions of language.

3.2 Functions of Language – A Cognitive-Sociolinguistic Perspective As opposed to the theories presented above, Kristiansen and Dirven (2010) re- frain from distinguishing between different language functions; rather, they turn their attention to the question of how linguistic items evoke particular stereotypes. It is claimed that the conceptual metaphor and metonymy theory, which was first presented by George Lakoff and Mark Johnson in 1980, may facilitate a

2 The production of an LL sign might be divided over by several individuals. Thus, the term sign-producer as opposed to sign-writer will be used in order to do justice to the several actors possibly involved in the production of a sign. 3 For a more detailed illustration of this understanding of context cf. Bublitz 2006. 9 deeper understanding of language in its function as a cultural symbol. Hence, it seems appropriate to introduce the basic notions of their theory. Challenging the original understanding of metaphor, which is traditionally understood as “a figure of speech in which one thing is compared with another by saying that one is the other” (Kövecses 2010: ix), it is argued that metaphoric concepts structure not only language, but also thoughts. Similarly, the nature of metonymy is believed to be conceptual (Kövecses 2010; Lakoff/Johnsen 2003), helping to understand abstract matters in everyday life. Simple concepts provide access to related, more complex domains of experience by making them more concrete. Relating to the suggested terminology, the more graspable source domain is used to concretise the abstract target domain (Lakoff/Johnsen 2003). While metaphor involves two completely separate domains, conceptual metonymy is believed to operate within a single conceptual structure (cf. Kövecses 2010: 108), where source domain and target domain are part of a single domain. To illustrate, in “metonymy we use one entity, or thing (such as Shakespeare […]), to indicate, or to provide mental access to, another entity (such as one of Shakespeare’s works […])” (Kövecses 2010: 172).4 This idea makes it particularly suitable for the present context. With regards to language use in the LL, a consideration of cognitive processes can help to understand the “differentiated conceptual links between language and culture” (Kristiansen/Dirven 2008: 2). In some cases, sign-producers make deliberate linguistic choices with the aim of evoking particular cultural stereotypes through language. Metonymic links connect individual lexical items with the languages they are thought to form part of. Subsequently, this is likely to create associations with countries, cultures, values and abilities that are conceptual- ly linked with the language used. More particularly, the presence of an individual lexical item or script system triggers the image of the whole language; these connections, in turn, reinforce associations with the respective culture. In the present paper, it is not maintained that the conceptual metonymy theory excludes the approaches presented above. Rather, it is meant to support these explanations and allow for a deeper understanding of the cognitive processes at work, providing a scientifically sound theoretical background for the argumen- tation.

4 Kövecses explains that there are several types of conceptual metonymy (c.f. Kövecses 173). PART FOR WHOLE metonymies are those cases of metonymy that are particularly interesting for the present paper (cf. Lakoff/Johnson 2003: 36, Kövecses 2010: 173). 10

The theories and concepts presented will now be applied to language use in the investigated LL, providing a scientifically sound basis for the methodological framework developed.

4 Reading the Curry Mile – A Case Study In order to explore the versatile functions of language and script visible in the public sphere, a case study on language use in the LL of the Curry Mile of Man- chester, UK was conducted. The next sections present the research objective, methodologies and a brief definition of the research area. Following this, it will be elaborated on the findings gained from the analysis of the collected data.

4.1 Research Objective

The dynamic and subjective character of context and meaning-making leads to the assumption that the functions of particular languages in the LL are not fixed or given beforehand. The questions of whether language on a sign actually communicates referential content, or whether it has a primarily emblematic or otherwise symbolic value is highly dependent on the individual reader. It is the purpose of the present study to explore the functions of different languages in distinct domains and contexts. The research aims at uncovering the potential motivations behind language or script choices, and it seeks to investigate the pragmatic and social functions of language possibly ascribed by individual readers. These issues will be addressed in several stages, and the linguistic theories illustrated in the chapters above will help to explore and understand the multiple functions of language in the LL. The analysis and interpretation of the collected data aims to answer the following research questions:

1 Which languages and scripts are displayed in the LL of the Curry Mile, and to what extent does the LL reflect the area’s linguistic composition and ethno - linguistic vitality? 2 Can patterns of language choice be linked to the types of sign-producer and the domains? How does this relate to the sign’s location outside or inside establishments? 3 In the LL of the Curry Mile, do certain languages function mainly as means of communication, while others are primarily used for symbolic functions?

With these questions in mind, it shall be possible to draw more general conclusions regarding language functions and their presence in the public sphere.

4.2 Methodological Considerations

For a comprehensive investigation of the LL, a detailed and contextualized analysis is necessary. The methodology followed in this paper combines a quantitative and qualitative approach, facilitating an exploration of the meanings behind language choices shaping the LL. The creation of a systematically collected and defined corpus allows to analyse language use patterns with respect to the quantitative strength of the languages and scripts, combinations and their distributions across domains and sectors. The data and conclusions drawn from the preliminary macro-level analysis will then serve as a basis for a detailed investigation of individual signs and establishments. A pilot study was conducted to test the effectiveness of previous methodologies and proved to be of major significance for the creation of the framework underlying the actual study. The following sections will elaborate on the items included in the corpus and present further methodological issues.

4.2.1 Survey Items

In order to obtain an overall view of language use in distinct areas and by a variety of LL actors, it was decided to include any kind of stationary written text found in the investigated area.5 Not only traffic signs, shop signs and billboards, but also temporary items like job advertisements became part of the corpus. Fur- thermore, stickers and graffiti are included in the analysis, as they are typically not controlled or influenced by any regulations and therefore „indicative of a community’s vernacular literacy practices” (Pennycook 2009: 20). Furthermore, the present study considers also the proper names displayed on the signs. Although business names are not necessarily intended to communicate referential content, they contribute to the linguistic composition of an LL and can provide valuable insight into the functions of languages on signs. The definition of LL underlying the present analysis does not include spoken language use in the public sphere. Similarly, it was decided to exclude non- stationary signs from the analysis. Indeed, it is important to acknowledge the un- stable and dynamic character of the LL, which is subject to constant change caused by a variety of LL actors. Yet, I believe that language use patterns on signs must be regarded in their relation to the contemporary urban space and the presence of other signs in their direct surrounding; regarding non-stationary texts

5 Being slightly more general than the term sign-producer, LL actor similarly refers to “actors who concretely participate in the shaping of LL by ordering from others or building by themselves LL elements according to preferential tendencies, deliberate choices or policies” (Ben-Rafael et. al 2006: 27). 12 in the analysis, it seems practically impossible to relate the sign’s linguistic features to any physical context. Turning attention to LL actors, it seemed necessary to distinguish between three types of sign-producers. Accordingly, signs issued by national and municipal bureaucracies, administrations and institutions were classified as “top-down” signs, whereas private and mostly commercially-driven signs fell under the category “bottom-up” signs. Figure 1a. on the left shows one of the top-down signs found, while Figure b. is an example for bottom-up signage.

Figure 1a. Top-down sign

Figure 1b. Bottom-down sign

The category of “bottom-up” signs included both LL items belonging to businesses located along the Curry Mile, as well as posters and signs advertising for non- local companies or institutions.

Finally, instances of mostly unauthorised and privately produced signs were cat- egorised as “transgressive” items (cf. Figure 1c. and Figure 1d.).6 Being more exact than previous frameworks, this methodology aims at capturing the varied forms and features of written text found in the LL of the Curry Mile.

Figure 1c. Transgressive item 1 Figure 1d. Transgressive item 2

Regarding the unit of analysis, different approaches were combined and the collected data was analysed in several steps. First, all signs were counted and analysed individually, regardless of whether they belong to the signage of an establishment or not. In a second step, particular attention was paid to the bottom- up corpus. LL items belonging to one establishment must be related to each other, since they eventually constitute an entity. Thus, the signage of each business was analysed and interpreted as an entirety. In order to gain a better understanding of the possible language functions on given signs, two defining principles are central to the present approach. First, it is believed that LL items must be investigated in their context of occurrence. Lin- guistic choices visible on bottom-up signs were correlated to commercial sectors and types of establishment. This domain-related approach shall help to investigate whether the use of minority languages is restricted to specific areas and products typical for the respective cultures, or whether they extend to further areas. Second, the investigation of language inside the local businesses was a significant part of the analysis.7 As mentioned above, it has become apparent in the pilot study that written language use inside might crucially differ from the linguistic characteristics of external signage belonging to the same business.

6 It must be noted that in some contexts, graffiti might be legal, and such instances could not be counted as “transgressive” text. In the investigated area, however, the graffiti found were clearly illegal. 7 Due to limitations in length, signs inside the establishments located in the research area were not counted and analysed quantitatively. 14

Since LL items outside are generally accessible to a wider and possibly more diverse audience than signs displayed inside an establishment, I assume in this paper that the semiotic functions of the respective texts may differ. The findings from these investigations then serve as a basis for the in-depth analyses of individual signs in their environment. The framework underlying the investigation of the individual signs is outlined in the following chapter.

4.2.2 Internal Variables of Signs

In order to facilitate a comprehensive and systematic analysis, the framework is structured into several categories and subcategories. The elements of classification are outlined below.

1. Linguistic Form 1.1 Language choice and combination 1.2 Presence of script systems on the sign 1.3 Remarkable linguistic features (unconventional use of language on orthographic, syntactic, lexical levels) 2. Language-Content Relationship 2.1 Informational value of each language used 2.2 Presence of proper names 3. Visual Features 3.1 Language-spatial relationships 3.2 Visual dominance (positioning, font size, typography, colouring, etc. 3.2 Multimodal aspects: meaning-carrying pictures, icons, symbols

As apparent from the outline, the framework integrates features related to the linguistic form of the verbal text, its denotational content and extra-linguistic aspects. In a first step, attention was paid to the characteristics of the verbal text displayed on each sign. Using a classification system suggested by Haarmann (1989), it was noted whether a sign was monolingual, bilingual or multilingual. Accordingly, a sign was considered multilingual if it contained at least 3 different languages or script systems (cf. Haarmann 1989: 55f.). Apart from an analysis of language and script, a consideration of combination patterns was essential in determining the functions of individual codes in creating meaning. Furthermore, language use was analysed in detail and striking orthographic or grammatical features were taken into account.

Second, the ideational content of the linguistic items is considered a fundamental aspect in the present paper. Different types of information might be given in distinct languages, possibly complementing each other to deliver a message. Reh (2004) suggests a method to analyse LL items with respect to information arrangement, which was applied in the present study. Reh distinguishes between four main combinations of languages and content, which she terms duplicating, fragmentary, overlapping and complementary (cf. 2004: 8). In duplicating multilingual writing, equal information is given in several languages; this combination of language and content is regarded as beneficial for monolingual readers who understand only one of the languages. Figure 2 below is an example of such duplicating multilingual writing.

Figure 2. Duplicating multilingualism

With respect to the signs’ symbolic message, Reh claims that duplicating writing “signals equality of all the linguistic and cultural communities thus addressed” (Reh: 2004: 8). It is important to emphasise, however, that aspects such as positioning, colour and font size may attach a language more or less dominance (cf. Kress/Van Leeuwen 2006). In fragmentary multilingualism “full information is given only in one language, but selected parts have been translated into an additional language or additional languages” (Reh 2004: 10). Overlapping multilingual writing appears in two types: frequently, two or more languages convey partially the same content but give additional information each. In the second type, the overall message conveyed in different languages may be identical, while the pragmatic form and hence “their interpersonal meaning is not” (Reh 2004: 12). 16

An example of overlapping multilingualism is given in Figure 4 below. Here, the English-speaking audience is requested to leave their umbrellas where the arrow points to; the Chinese version of the text, on the other hand, is a general state- ment and can be translated as ‘There is an umbrella stand’.

Figure 4. Overlapping multilingualism Complementary writing responds to individual multilingualism and presup- poses a multilingual readership. These are texts in which distinct parts of the overall information are each given in a different language (cf. Figure 5 below).

17 Figure 5. Complementary multilingualism While the Arabic text translates as ‘special sale offer’, ‘whole lamb’, the English text offers information about the dates of the offer and the discount. Since “knowledge of all the languages involved is required to understand the whole message”, Reh hypothesises that the sign-writer expects readers to understand each of the languages used (cf. Reh 2004: 14). In the examination of translation types, Reh pays attention to the relationship of co-occurring languages only; their relative informational value as such is not taken into account. The present study, in contrast, investigates the semiotic values of the different parts and takes the functional relations between the different languages into account. Apart from the linguistic features and the communicative force of different codes, their relative visual dominance was investigated. Communication occurs through multiple modes interacting to create meaning, which is why the consideration of multimodal ways of meaning-making is of crucial importance.8 The analysis of visual aspects of signage draws on the work of Kress and Van Leeuwen (2006) and considers spatial relations, salience and the presence of pictures, icons or illustrations. The versatile framework established for the present study provides a much needed qualitative perspective, helping to identify the possible communicative and social functions of language on LL signs. The following section defines the area investigated for the present LL study and briefly describes the fieldwork. Providing background information about the research area, its ethnolinuistic pro- file and language policy helps to establish the relevant context of interpretation.

4.3 Research Area and Fieldwork – The Curry Mile

The fieldwork for the present study was carried out from October 2013 to Febru- ary 2014 and consisted of systematically collecting all stationary texts visible along a 1 km-long stretch of Manchester’s Wilmslow Road.9 Famous for its South Asian and Middle Eastern gastronomy businesses and grocery shops, this part of the street is commonly known as the Curry Mile (cf. Figure 6 below.

8 Cf. Hoffmann (2010: 10ff.) for a more elaborated illustration of multimodal meaning- making. 9 In the qualitative analysis, a few individual signs found in the surrounding area will be included for their interesting linguistic choices. 18

Figure 6. The Curry Mile, Manchester, UK. (Google Maps. URL: https://www.google.co.uk/maps/@53.4537983,-2.2251743,15z. last accessed February 18, 2014).

The Curry Mile is located in Manchester’s suburb Rusholme, an area characterized by a significantly diverse ethno-linguistic composition.10 A closer look at the Rusholme census (cf. NOMIS 2011) reveals that less than 40% of the population are white British; astonishingly, a similar proportion of the population have South- Asian background. According to the census, approximately 71% of the adult population speaks English as their main language, while almost 15% indicate their main languages to be a South Asian language like Urdu, Bengali or Panjabi (NOMIS 2011a, NOMIS 2011b).11 An essential aspect in the research of signs is the consideration of language policies, which may influence the LL. England does not have any constitutionally defined official language, and there is no legislation in the UK that would make the use of English or any other language on public signage mandatory (cf. Dona- key 2007). Yet, English is obviously spoken and written as a main language by the vast majority of the country’s population, and it is functionally specified as the language used in formal education, media and business.

10 It is necessary to note that therefore, the investigated LL is not representative of the city of Manchester as a whole. 11 Indeed, an interpretation based on census data must be done with care, as indications made may be inaccurate. 19

Turning back to the analysis of the LL of the Curry Mile, each local establishment was in a preliminary step listed according to its type in order to maintain an overview over the great number of LL items and to facilitate a comprehensive analysis. During the actual fieldwork, each sign was photographed and notes on language use were taken. Since the methodological approach developed for this study considers also written language use inside the establishments, posters and signs in the shops, agencies and offices were investigated and, if permitted, pho- tographically documented. In the next step of research, the collected corpus was subjected to a comprehensive analysis following the framework presented above (cf. Figure 1). In order to ensure the accuracy of the analysis, all signs were re- viewed in the presence of academics who have good command of the respective languages.12 In the following sections, I will present the findings of the empirical study conducted for the present paper.

5 Language Use on Signs in the Curry Mile – Quantitative Findings In the research conducted for the present study, 781 individual units could be distinguished.13 In a first step of analysis, the collected data was classified according to signs-producers. Figures 7a and 7b below illustrate the proportion of top down, bottom-up and transgressive LL items found in the investigated area.

Proportion LL Actors

Agency n % Top-down 104 13,32 Bottom-up 629 80,54

Top-Down Bottom-Up Transgressive 48 6,15

Transgressive Total 781 100

Fig. 7a. Proportion LL Actors – Pie Chart Fig. 7b. Proportion LL Actors

At 80%, commercial and other kinds of bottom-up signs make up the numeri- cal majority; in contrast, only 13% of texts were counted amongst the top-down corpus.

12 Cordial thanks to Prof Yaron Matras (University of Manchester), who helped with the translation of some texts. Similarly, Serena Fragagnano, Ru Lan and Darya Gorbunova from the Manchester Grammar School deserve special thanks for translating Italian, Chi- nese and Russian signs, respectively. 13 The entire corpus data can be found on the electronic storage medium enclosed. 20

With only 6%, transgressive items take up the significantly smallest part of the LL of the Curry Mile. Turning to the degree of linguistic diversity, combination of languages and their functions, it becomes evident that the three categories exhibit extremely different characteristics (cf. Figure 8 below).

Top-Down n (%) Bottom-Up n (%) Transgressive n (%) Monolingual 101 (97,11%) 478 (ca. 76%) 48 (100 %) Multilingual 3 (2,99%) 151 (14%) 0 (0%)

Figure 8. Proportion Monolingual and Multilingual Signs within Category.

Within their category of ‘bottom-up’ signs, the proportional share of multilingual signs is considerably higher than in the other two samples. In the categories of top-down and transgressive items, monolingual signs are clearly predominant; strikingly, none of the transgressive items contains more than one language. Furthermore, an analysis of the data has revealed that the maximum number of codes combined on individual pieces of LL signage is for bottom-up signs higher than for the others. Likewise, the range of distinct language present in the former category is much more diverse than the linguistic variety of bottom-up and transgressive items. 14 different languages are used on bottom-up items, as opposed to 5 languages in top-down signs and 2 on transgressive items. Laying the basis for qualitative analyses, a more detailed investigation of the patterns of language use in these three categories is illustrated in the following.

5.1 Top-Down Items

The first category to be analysed was the sample of governmental or otherwise official signs found in the LL of the Curry Mile. As illustrated above, an over- whelming majority of signs of this group is monolingual. Reflecting its central significance in official discourse in the UK, English is contained on each of the signs. Less than 7 % of the top-down signs visible in the area contain a language other than English, displaying minority languages such as Arabic and Urdu. No- ticeably absent from the top-down signs displayed along the Curry Mile are Kurd- ish, Farsi and Bengali, which are all strongly represented in the area of Rusholme (NOMIS 2011b). Linguistically less diverse than top-down signs are transgressive items, an analysis of which will be illustrated in the following.

5.2 Transgressive Items All unauthorised texts found in the investigated area are of monolingual character, and more than 95% are English. Regarding graffiti, the task of determining which language is used was more difficult. Since many consisted of short tags or acronyms, it was in 3 cases impossible to unambiguously determine the language. From 9 different graffiti that could be distinguished, 5 are to be catego- rised as monolingual English. Only one item of this category is not English, conveying only Chinese. Much more linguistically diverse is the bottom-up corpus, which will be illustrated in the following.

5.3 Bottom-Up Items One of the first quantitative questions posed was concerned with the frequency of present languages and script systems. The languages used are listed and ranked according to their frequency of appearance in Figure 9a and in 9b below.14

Language Contained on n of signs Contained on % of signs 1. English 603 68,76 2. Arabic 151 17,22 3. Urdu 50 5,70 4. Kurdish 22 2,51 5. Chinese 18 2,05 6. Italian 9 1,03 7. Farsi 8 0,91 8. Hindi 4 0,46 9. Somali 3 0,34 10. French 3 0,34 11. Spanish 2 0,23 12. Russian 1 0,11 13. Malay 1 0,11 14. Turkish 1 0,11

Figure 9a. Languages According to Frequency of Appearance

14 An overview of the entire bottom-up corpus, which lists all establishments and their signage, offers Figure 30 (on electronic storage). 22

Languages Contained on Signs

700

600

500

400

300

Numberofsigns 200

100

Language

Figure 9b. Languages According to Frequency of Appearance – Bar Chart

It becomes apparent that the three languages most strongly represented are English, Arabic and Urdu, which reflects the nature of the ethno-linguistic make- up of Rusholme and coincides with language use in the top-down sample. Unsurprisingly, English is also in this corpus the most dominant language. All the more striking is the strong presence of other languages such as Arabic, which was recorded in almost 20% of the corpus. Appearing on 50 signs, Urdu is the third most frequent language among bottom-up signs. The remaining languages appear on less than 5% of the signs, with Turkish, Malay and Russian conveyed on only one LL item each. An analysis of script systems present in bottom-up signs reveals that five different script systems are used, appearing in a variety of combinations. Figure 10 reveals that almost one third of the monolingual and multilingual bottom-up signs display information in a script other than the Roman alphabet.

Script System Present Number in n Proportion in % Latin script 491 71,16% Arabic script 176 25,51% Chinese Characters 18 2,61% Devanagari script 4 0,58% Cyrillic script 1 0,14% Figure 10. Presence of Script Systems

The Arabic script is present on 25% of the signs, which makes it the second most frequent writing system in the LL of the Curry Mile.15 It has already been mentioned that English monolingual signs make up a large part of the entire LL. Of specific interest is, however, that there are 65 monolingual bottom-up signs using a language other than English. Almost 25% of these are monolingual Arabic, 16% contain Urdu and a similar proportion of signs is monolingual Chinese. While Italian is rarely used monolingually, French is not assigned such an autonomous role at all. As much as in monolingual signs, Eng- lish has also in multilingual signs the highest frequency of occurrence. In fact, one of the key outcomes of the analysis is that all of the multilingual signs contain English, indicating that the language is intelligible to wide audiences and intended to make communication between various language groups possible. It is by far mostly combined with Arabic, making up approximately 65% of bilingual signs and posters. Two combination patterns that were also quite strongly represented are English-Kurdish and English-Urdu. A further noteworthy point is that Kurdish never appears autonomously, but only in combination with English. This implies that it is for a large part of the potential audience not considered an autonomous means of communication (cf. Bagna/Barni 2006). Having obtained a general picture of the presence and the combination patterns of languages and scripts present in bottom-up signs in the investigated area, it is now interesting to turn to a domain-related analysis of this part of the corpus.

15 For a better understanding of the analysis, one must note that the Arabic script represents more than just spoken Arabic. Its modified versions are in the investigated LL used to write Urdu, Kurdish and Farsi. 24

5.3.1 Language Use across Domains

The analyses of the patterns of language use related to the different domains have revealed insightful results (cf. Figure 11 in the Appendix). First, it is striking to note that a considerable number of sectors use exclusively English on their sigs, whereas signage of other domains is characterized by multilingualism or even predominance of languages other than English. Businesses of the automotive sector, jewellers, convenience stores, solicitors and other services display information only in English. With an insignificant number of exceptions, this holds also for private medical institutions and the finances sector. Likewise, businesses falling under the category “PC & Electronics” are marked by a clear preference for monolingual English signs (almost 90%). Signage belonging to restaurants, on the other hand, is characterised by bilingual patterns; here, the combination of English and Arabic is predominant. In this category, almost 30% of signs contain Arabic. In comparison, just over 50% are monolingual English. Furthermore, English is typically used alongside with other South Asian, East Asian or European languages. The gastronomy sector has, interestingly, the highest percentage of Chinese, whereas this language is absent in practically all of the residual domains. As mentioned above, an investigation of language use inside the establishments may provide valuable insight into functions of language on external signs. In the first part of the following section, an overview of language use outside and inside all the establishments along the Curry Mile will be taken. In relation to the results presented above, this creates an indispensable basis for initial interpretations.

5.3.2 Language Use Outside and Inside Establishments

In the analysis conducted for the present research, some major discrepancies between language use on external signage and code choice on signs displayed inside a given establishment have been detected.16 From the total of 187 stores, restaurants, shops, offices, institutions and agencies located along the Curry Mile, more than a third exhibit differing code choices for external signage and internal signs. The most common pattern is the use of bilingual English-Arabic or English-Farsi signs outside, while signage inside is monolingual English (15% of all establishments). Another interesting result is that 100% of the premises in the

16 A detailed overview can be found in Figure 12 on the electronic storage medium. 25 area that use either Italian or French outside do not use it for more detailed information inside. In 10% of all establishments, on the other hand, the exclusive use of English on signs visible from the street stands in contrast to the multilingual signage inside. In these cases, the community languages Arabic, Urdu or Kurdish are additionally used inside to communicate information about products and services.

5.4 An Initial interpretation

Having analysed the entire corpus, an interrelation of the findings allows for an initial interpretation with respect to the research questions posed above. Some languages are not present at all, although the census suggests that a considerable number of speakers are local to the area (NOMIS 2011b). Speakers of Guja- rati and some African languages are quite strongly represented in in the area, but the languages are absent in the LL. Urdu seems to be disproportionately un- derrepresented if one considers the large proportion of Urdu-speakers in the area’s population. Furthermore, there are more Romanian- and Russian-speakers than individuals who regard Italian or Spanish as their main languages (cf. NO- MIS 2011). Hence, the presence of Italian, French or Spanish in the LL of Rush- olme is relatively striking considering that there are hardly any speakers of those languages living in the area. Comparing code choices across authorship and domains, it becomes apparent that the three categories of sign-producers show distinct characteristics of language use. While English is predominant in the entire LL, bottom-up signs are linguistically more diverse than the other types of signs. The omnipresence of English suggests that the language functions as lingua franca in the LL of the Curry Mile, facilitating communication across communities. As a consequence, the question arises whether languages other than English are actually necessary as a means of communication in the investigated area. These findings lead to a number of assumptions regarding the motivations of different LL-authors influencing code choices for signs. One can assume that certain language functions are more likely to be prominent on commercial signs, while others seem to be characteristic of official signs. With respect to top-down signs, it could be argued that referential and conative functions are foregrounded. Identifying places or informing about restrictions, warnings or dangers, their main purpose concentrate on the communication of information and the regulation of the addressee’s behaviour as opposed to expressing feelings or exerting appealing effects on the reader.

To ensure understandability of essential information, warnings or bans, language choice is most likely to be oriented towards the expected linguistic competencies of the readers (cf. Spolsky/ Cooper 1991: 89). Figure 13 below serves to substantiate this claim.

Figure 13. Bilingual Top-Down Sign “Islamic Centre”

The sign pointing to the Manchester Islamic Centre is a bilingual top-down sign found on Oxford Road, the extension of the Curry Mile towards the city centre of Manchester. The choice of Arabic in addition to English indicates that linguistic choices were made according to the assumed linguistic proficiencies of those readers who are most likely to require the information given. Since the sign underneath is likely to be most relevant to an Anglophone audience, the monolingual English sign implies that the local government adapts to the assumed linguistic competencies of the target readership. The dominance of English on transgressive items suggests the importance of the language for individuals. Language choices are practically unregulated and indicate language knowledge and attitudes. Turning to bottom-up signs, it is reasonable to state that motivations other than the mere communication of information crucially influence code choices. Although commercial signs may very well be intended to serve an informative function, the denotational content of linguistic items itself is not necessarily the primary message to be delivered. Thus, it can be claimed that language choice does not always comply with the sign- writers’ linguistic competencies, nor with the assumed readers’ language knowledge.

These hypotheses suggest that the linguistic diversity in the collected bottom-up data can at least partly be explained by the variety of reasons motivating linguistic choices (cf. qualitative analysis below). Returning to an overall perspective on the entire LL, the data has revealed that languages cluster around particular domains, which might be related to the connotative values of languages. Besides the fact that code choices seem to be influenced by domains, the findings indicate that language use is crucially dependent on where the sign is located and to whom it is accessible. In fact, it seems that there is a strong relationship between the distribution across sectors and the findings related to language use on outdoors signage versus interior signs. In most cases, striking results of the investigation of language use inside correlate with the findings that emerged from the domain-related analysis. In those domains where certain languages are likely to carry symbolic values as they can be associated directly with the respective products, the discrepancies between linguistic choices inside and outside were remarkably consistent. More precisely, minority languages frequently appear on external signage of particular gastronomy businesses, while they do not seem to be used as a main means of communication inside. On the other hand, English is prevailing on the façades of those businesses where its connotations of reliability, progress or international orientation seem to be profitable (cf. Haarmann 1989). The correlation of the results from the domain-related analysis and the investigation focusing on language use outside versus inside the establishments serve as a basis for the qualitative analysis presented in the following. The more detailed investigation will further elaborate on the hypotheses formulated above. Moreover, it will be shown that languages may serve various functions, and that distinct motivations may determine the sign-producers’ linguistic choices.

6 Language Functions in the LL – A Qualitative Perspective

The following chapter presents a number of revealing analyses that all aim at answering the research questions presented above. Striking results from the quantitative analysis comparing written language use outside to inside are taken as a starting point, being regarded as essential indicators for the roles that certain languages might play in a particular context.

6.1 Language in the LL - An Expression of Identity and Authenticity

The sign depicted in Figure 14. illustrates that different scripts and languages may fulfil different roles in the meaning-making process.

Figure 14. Wanasah Café

It belongs to the shisha bar “Wanasah Café”, which displays four monolingual English signs and two bilingual English-Arabic LL items outside. Written language use on menus inside, on the other hand, is restricted to English. The sign (cf. Figure 14) displays the café’s name in both Latin and Arabic script, but the products listed underneath are given in English only. The type of functional arrangement (Reh 2004) is thus fragmentary, denying monolingual Arabic-speakers part of the information. With respect to content as such, it seems reasonable to claim that the English part is of greater informational value. Taking a multimodal perspective, it could be argued that the illustrations accompanying the text inform non-English speakers about part of the product range. Yet, the pictures are no adequate and comprehensive visual translations of the content given in English. Likewise, the bilingual sign belonging to “Stars Shisha Café” (cf. Figure 15.) combines English and Arabic in a fragmentary way. Only the proper name is given in both scripts, but the list of products and the visually dominant writing LIVE FOOTBALL is not repeated in Arabic.

The accompanying pictures and illustrations could be useful for monolingual Arabic-speakers, but it seems reasonable to claim that they are intended for aesthetic enhancement rather than to enable potential non-speakers of Eng- lish to understand the verbal text. As a close linguistic analysis of the Arabic version of the text Stars Shisha Café reveals, it is a clear result of language contact. The Arabic part is merely a translitera- Figure 15. Stars Shisha Café tion of the English lexemes into the Ar- abic script, in which ‘shisha’ and ‘café’ are not actually translated into Arabic. For monolingual readers of Arabic, the referential content would not be comprehensible.17 Strikingly, also the word order follows English syntax. An analysis of language use inside shows that information on products is given only in English, while written forms of Arabic are entirely absent. In an overall perspective, it can be claimed that the language functioning as communication tool is in both cases English rather than Arabic. As in a great number of establishments in the investigated area, the way in which external signage differs from language use inside suggests that English is the primary communication tool. The referential function of the Arabic text is likely to be subordinate to its emotive, conative, phatic and aesthetic functions of language. First, the linguistic items serve a clearly expressive function, communicating information about the sign-writer and expressing identity. The presence of Arabic script may have an appealing effect to Arabic-speaking readers, demonstrating solidarity and creating a certain in-group-feeling (Gumperz 1982: 66). This can be linked also to the phatic function of language, which relates to the potential of language to establish a communicative and social contact. Particularly from a European perspective, the Arabic script is likely to enhance the sign’s ability to draw attention.

17 It could be claimed that not only Stars is to be regarded as the café’s proper name, but the entire phrase Stars Shisha Café. This would account for the fact that the lexemes are not translated, but only transliterated into the Arabic script. However, the use of the Ara- bic script as such still indicates its ornamental function. 30

As it contrasts the unmarked Latin script dominating most LL items in the investigated LL, the presence of Arabic makes the sign more conspicuous and may attract the passers-by’s attention. Additionally, the script may acquire iconic power, representing the Arabic culture and associated competencies. Seargeant notes that a script system is often considered “as part of a particular ‘language’” (2012: 187), which influences how a sign operates as a communicative act in a particular context. Even though readers may not be able to understand the actual message, they will most likely recognize the script, leading to metonymic links to the Arabic culture. These assumptions support the abovementioned claims that language- choice on commercial signs can be related to the products and services advertised (cf. Kelly-Holmes 2005). Arabic-speaking countries may be regarded as those world regions that are, among others, commonly associated with the cus- tom of smoking shisha. Hence, the use of Arabic script on the signage of shisha bars can be linked to the country-of-origin effect (cf. Kelly-Holmes 2005: 67). From a cognitive-sociolinguistic perspective, it can be argued that the Arabic script evokes metonymic links to particular values associated with the respective culture, which is likely to add to the products’ authenticity. In conclusion, it has become evident that languages may be used for their power to attract, to establish identity or create connotations rather than for their function as communicative tools. The example presented in the following section suggests that in some cases, linguistic choices may sometimes be entirely de- tached of ethnicity and language knowledge of the participants, but closely related to the offered products.

6.2 Commodification of Language and Detachment of Ethnicity In previous chapters, it has been maintained that there is a correlation between the domain-related analysis and the extent to which external signs differs from language use inside. Codes whose use is concentrated on particular sectors appear in these sectors mainly on the façade of the businesses, but they are not actually used for interaction inside. This also applies to the use of Italian and French in the investigated area, which are chosen only for external signs of the gastronomy or the fashion sector respectively. Strikingly, census data suggests that not many people inhabiting the area speak either of the languages as their main language, which is another indication for the codes not functioning as actual means of communication in the present context.

Not far from the Curry Mile, the signage of a café whose façade can be seen in Figure 16 allows for interesting conclusions. The use of Italian is restricted to the café’s name, which is visually emphasised through positioning, colouring and font size. English is assigned a more significant role in communicating relevant information, as it is used to inform about the product range and whether the café is opened or not. Figure 16. Caffè La Fresco

The fact that the welcome sign is English is a further indicator for the expected clientele being Anglophone. It establishes a contact between sign-producers and the English-speaking audience. Returning to the Italian part, the ideational meaning of the café’s name may indeed be intelligible for part of the audience and connected to the freshness of the products offered inside; yet, it seems reasonable to claim that the association evoked by the language as such is held to be more important than its denotation. Kelly-Holmes argues that in some cases, “the communicative or utility value of the particular words has come to be obscured or mystified through the process of fetishization to the point where it becomes irrelevant” (2005: 24). Additionally, it is fundamental to note that caffè la fresco is most likely not worded by a Native speaker of Italian, as it is morphologically and syntactically no correct use of the language.18 Just as the country-of-origin effect is applicable to explain the language choice for the café’s signage, the cognitive metonymy theory serves to provide an understanding of the linguistic choices from a different perspective. Italian seems to be chosen in order to attach the offered products certain values that are frequently associated with the culture. The Italian name could be seen as the source domain. It is conceptualised as forming part of the Italian language and leads the reader to the respective fame, triggering associations with particular cultural stereotypes (target domain). The use of Italian can be regarded as a

18 Cordial thanks to S. Fragagnano for the confirmation of this claim. 32 cognitive reference point which triggers metonymic links to certain values and creates positive associations about food (cf. Piller 2003: 173 f.; Kristian- sen/Dirven 2008). In the examples illustrated above, Arabic and Italian are only part of the businesses names, which indicates that the ideational content is subordinated to emotive and affective meanings. It seems inappropriate to define the actual vitality of languages in a given area simply by quantitatively counting and determining languages displayed on the signs. The previous chapters have illustrated signage of establishments that use several languages on sign on the façades, but only English for more detailed information inside. The following section, in contrast, presents examples where signs inside are linguistically more diverse than external signs.

6.3 Monolingualism on External Signage

A recurrent pattern of language use was found in the medical and the financial sectors, as well as businesses falling under the category “PC & Electronics” or “Travel & Cargo” (cf. Figure 12). Considering the fact that all of them display monolingual English signs on their façades, it is interesting to note the additional use of minority languages inside many of the establishments. The signs displayed inside “Sylhet Travels”, for instance, show that Bengali is chosen to inform about the services and even used on regulatory signs (Figure 17). Strikingly, official information is given in English only.

Figure 17. Sylhet Travels

Turning to businesses of the financial sector, the international company “Western Union” is a revealing example of monolingual English signage on the façade, whereas additional languages are used indoors. The fact that services and offers are presented in Urdu inside indicates the language knowledge of at least part of the audience (cf. Figure 18.). 33

Translated, it means ‘welcome - good news – we are offering free bank pay- outs for our customers using UBL’. Considering the fact that UBL is a Pa- kistani bank institution, language choice implies the target-readership and specifically addresses this particular audience. It is also striking to note the fact that the note is handwritten, implying that the poster is not designed central- ly by the company.

Figure 18. Western Union

Rather, it seems that employees working in the branch have adapted advertising to the linguistic competencies and needs of the expected customers in this particular area. Furthermore, the handwritten note itself can be regarded as an indication of the sign-writer’s proficiency of Urdu.

In conclusion, the exclusive use of English in the abovementioned domains is indicative of its symbolic meaning, being commonly associated with reliability or future orientation. Its use on the façades of travel agencies can further be connected to its reputation as “the language of tourism and international travel” (Cheshire/Moser 1994: 461).19 Additionally, it indicates that English is generally expected to be understood in the respective community, and using English on signs visible from the streets makes the more inclusive to wider audiences. The prominence of written Bengali, Urdu or other minority languages indoors indicates the need to sometimes offer complex information in a language other than English. Being aware of the linguistically diverse clientele, LL actors choose to adapt to the assumed language knowledge of the expected readership. The fact that these languages are not displayed outside, on the other hand, could be indicative of language attitudes. These findings further reinforce the claims that external signage cannot readily be regarded as an indicator of language vitality or the strength and presence of particular language groups in the area.

19 However, there is also a considerable number of travel agencies in the investigated area that display multilingual signs on their façades. 34

6.4 Minority Languages in the LL – A Means of Communication

Discrepancies between written language use outside and inside have illustrated that symbolic values of a language seem in some cases to be foregrounded to its role as a means of communication, which co-determines code choices. It is now interesting to explore cases in which a particular set of languages is used on signage both outside and inside in order to investigate whether there are connections to informational and symbolic usage of language. Figure 19 below shows part of the external signage of “Syria Sweets”, an analysis of which illustrates that Arabic and English are quantitatively, visually, as well as content-wise equally important. Apart from the bilingual main sign above the entrance, there are two LL items at the sides of the shop windows that refer to the car park in Arabic and English. While the use of Arabic on other signs could indeed be linked to the products offered by the shop, it is unlikely to use Arabic for symbolic reasons on regulatory signs.

Figure 19. Syria Sweets Outdoors 1

The main purpose here is to communicate factual information, which is then intended to lead to a particular reaction. Figure 20 below depicts another sign belonging to the same shop, which is characterised by overlapping multilingualism (Reh 2004). The shop name is displayed in both languages, while contact details are given in English only. More importantly, relevant information on the product range will be comprehended only by an Arabic-speaking readership.

The pictures accompanying the verbal text may indeed give an impression of the products, but an Anglo- phone readership is denied detailed information.

Figure 20. Syria Sweets Outdoors 2

An analysis of language choice inside the shop reinforces the assumptions that Arabic is for some of the expected readers a necessary means of communication, since product names and further information is consistently offered in Arabic and English (cf. Figure 21a. and Figure 21b. below).

Figure 21a. Syria Sweets Indoors Figure 21b. Syria Sweets Indoors

So far, the investigation has specifically addressed the first and the second research question. Both quantitative and qualitative analyses have served to draw conclusions regarding the composition of the LL of the Curry Mile and its relation to ethnolinguistic vitality; furthermore, it has provided revealing insight into language use across LL actors, domains and location. The following section is particularly targeted towards the third research question and focuses on the claim that the functions of language in the LL depend on subjective perception of individuals.

6.5 Language Functions and Reader-Dependency

The following chapter presents two striking examples serving to illustrate that languages cannot be assigned a particular set of values and functions.

Figure 22 shows a bilingual sign advertising the range of services of a cargo company. Not only the language-space relationship, but also the language-content relationship indicates that Arabic is the dominant code in relation to English. Information is presented in a complementary way (cf. Reh 2004), and for a full undestanding of the entire information, proficiency in both languages would be required.

Figure 22. Al Arabia Cargo Company

European readers who do not have a command of Arabic will only understand the bottom part of the poster, which gives contact details and refers to the company’s official registration. Apart from this, an English text printed in red refers to their alleged efficiency. It can be claimed that the code choices imply a bilingual readership, and it seems that the two languages are chosen for different reasons. The English text communicates rather official information and emphasises the company’s quality. Only the Arabic text, on the contrary, informs readers about the service range.20 Turning to the English contact details, it seems reasonable to present them in the language of the country the company is based in. Following Sebba (2012: 101), language choice could also be regarded as an instance of metaphorical use of languages. Accordingly, English “metaphorically indexes the here-and-now” in England, as opposed to the Arabic text referencing foreign countries. This, however, does not in itself account for advertising their efficiency in English; in fact, it leads to the assumption that values possibly associated with English have influenced code choices. Nevertheless, the presence of the languages is most likely to be perceived differently by different readers. For an Arabic-speaking readership, the use of their language may be particularly appealing and evoke a sense of familiarity and in-group-feeling. Additionally, the presence of English may create a sense of

20 Regarding visual context, the illustrations may help to understand part of the message. 37 reliability. European readers, on the other hand, may link the use of Arabic script to the offered services and make assumptions about the business focus and the linguistic competencies of business owners, as well as the intended clientele. Likewise, the signs displayed by the sweet shop “Delhi” suggest that the languages Urdu and English fulfil distinct functions. Considering the signage in an overall perspective, the latter is clearly dominant and suggests that the expected customers speak English. Only one LL item is a monolingual Urdu note attached to the shop window (Figure 23, left), which seems to target a different readership. Translated into English, the handwritten note means ‘we are looking for staff’. Since there is no sign that conveys an equivalent message in English, the job offer is specifically addressed to an Urdu- speaking audience.21

Figure 23. Delhi Sweet Shop

While the text conveys an ideational message to an Urdu-speaking readership, the referential functions will not be fulfilled for a readership lacking knowledge of the language. This, in turn, foregrounds other language functions and readers possibly ascribe meanings to the sign that were not intended. For a European audience, the script as such is likely to create metonymic links, leading to assumptions about the LL actors’ origin and the authenticity of the products. It becomes evident that the functions of language and constitution of identity through code-choice is crucially determined by the perception and knowledge of the audience. Viewers are likely to recognise the script-system as Arabic, but those who are unfamiliar with its usages will not be able to identify the language as being Urdu. Rather, the Arabic script may trigger associations other than those intended by a sign-producer who wants to express his identity.

[I]n order for an identity to be established, it has to be recognised by others. That means that a lot of what happens in the field of identity is done by others, not by oneself. (Blommaert et al. 2005: 205-206) For different readers, hence, language might perform distinct functions.

21 Presenting this information monolingually may have several reasons linked to the sign- writers language skills or the audience that is supposed to be addressed. With the limited scope of this paper, this cannot be illustrated here. 38

Regarding the producers of the signs, one can claim that they address different groups of readers in different languages, conveying distinct messages to different audiences. Comparable to the phenomenon of code-switching, deliberate language choices for LL signs might be used for addressee-specification, serving to “direct the message to one of several possible addressees” (Gumperz 1982: 22).

6.6 The Functions of English in an Anglophone Context

Of particular interest in an investigation of language use in the LL of an Anglo- phone country is the signage of the shop “Sweet & Salt”, an analysis of which suggests that minority languages are used as primary means of communication, while the functions of English are for some readers symbolic rather than informative (cf. Figure 24 below). On its façade, the shop uses Arabic, Farsi, Kurdish and English for a variety of purposes. Arabic script is visually dominant, and regarding the informational values of the codes used, English is in a subordinate position. The black-and- white sign above the window is an example of complementary multilingual writing (Reh 2004). The name of the establishment is given only in English, which is pre- sumably the only part of the sign accessible to most European readers.

Figure 24. Sweet & Salt Outdoors

The Arabic text written on the left of the shop, on the other hand, offers information about the products offered inside.22 The monolingual sign above the door displays ‘welcome’ in Arabic and is thus specifically appealing to Arabic- speakers. As on the main sign, the remaining LL items offer detailed product information mainly in languages other than English. The vertical sign is trilingual and conveys the Arabic, Farsi and Kurdish lexemes that can be translated as ‘snacks’. The illustration underneath takes up the general theme, but it does not visualize what the verbal texts refers to. In a general perspective, the linguistic choices limit accessibility to detailed information to an audience that reads Ara- bic, Farsi and Kurdish but excludes others. An investigation of language use inside reveals that all product information is given in English and Arabic (Figure 25).

Figure 25. Sweet & Salt Indoors

The analysis indicates that the different languages serve distinct functions. In contrast to the use of Arabic on the shisha bars’ signs, it is here intended to make information accessible to non-speakers of English. The Arabic ‘welcome’-sign further indicates the orientation towards an Arabic-speaking audience, for whom the sign may create a sense of in-group-membership and emotive connections. At the same time, the LL actors assert their own identity using Arabic as a “we- code” (Gumperz 1982: 66) and exhibiting particular commitment with this group. It opens a channel for communication, expresses a social relationship with the LL actors, suggests certain attitudes and creates a sense of familiarity. Furthermore, the data suggest that Arabic functions as a main means of communication, whereas English may even be used to trigger particular associations rather than to communicate content. For a speaker of Arabic, the English shop name may give the setting a prestigious touch and trigger associations with a fashionable westernized lifestyle. The alliteration in “Sweet & Salt” clearly foregrounds the aesthetic value and thus poetic function of language, enhancing its symbolic power. Furthermore, the language could have been chosen for more personal reasons connected to the sign-producer’s diverse linguistic identity of

22 The physical environment including the window display may be essential in meaning- making. Yet, material objects can in the present research not be taken into consideration. 40

(cf. Matras 2008). Using both English and minority languages, the LL actors establish connections to their new environment, as well as their home countries. This suggests the desire to be identified with both languages and cultures. It has been mentioned that Farsi and Kurdish are not used inside to facilitate the respective audiences an understanding of the product information. This further supports the claim that language use on external signage does not always actually “indicat[e] that the language in question can be used to communicate and obtain services […]” (Landry/Bourhis 1997: 25). Rather, Farsi and Kurdish seem to be chosen to address and attract the respective passers-by, signalling solidarity. It is highly expressive and reveals the aims of developing and main- taining social relations with these language groups.23 The final qualitative analysis presented in the following further illustrates that functions and symbolic values of language are highly subjective.

6.7 Language-Contact and its Functions

The sign in Figure 26. is a bilingual sign using both Urdu and English. It seems to be designed for a linguistically diverse readership, facilitating both Urdu- and English-speakers to understand at least part of the information given.

Figure 26. Special Lahori Paan

23 The fact that information is given in Arabic, Kurdish and Farsi further supports the claim that Arabic script is used not only for emblematic reasons. For readers unfamiliar with the script, the languages are hardly distinguishable from each other. If the Arabic script was chosen for its aesthetic value only, there would be no reason to write the messages in different languages that use the same script. 41

However, each of the languages is claimed to have an influence on either of the language groups when making meaning. First, the sign signals the sign-writer’s intention to reach wide audiences, possibly leading to assumptions about the ethno-linguistic composition of the customer-base. Furthermore, the Arabic script system, particularly for European readers, create metonymic links to the cultures and products associated with it. A closer investigation of the Urdu part reveals that for Urdu-speakers, English may serve symbolic functions. Being a case of code- and script-mixing, the text has a multilingual character in itself. The Urdu text printed in red literally say ‘special lahori paan’; to be more precise, the English lexeme special has simply been transliterated into Urdu script instead of using the actual Urdu term. Likewise, the Urdu translation of special lahori kulfi & falooda contains the English lexeme special represented in Arabic script. Finally, also the Urdu version for sweet corn is simply the English word written in Urdu script. Hence, it would not be of any informational value for monolingual speakers of Urdu. The referential function of the lexical items is fulfilled only for a bilingual audience who can read the Arabic script and understands English. Urdu-speakers who do not have sufficient knowledge of English can interpret special and sweet corn phonetically, but not semantically. Yet, it is likely that monolingual speakers of Urdu will recognize the suggested sounds as being English, which could have a symbolic meaning in itself. Returning to the conceptual metonymy theory, the evoked English sound (source domain) may be related to the English language, which is metonymically linked to the wider frame of particular cultural stereotypes such as a fashionable lifestyle (target domain). Even if the content is understood, the fact that the English lexeme was preferred to its Urdu translation may carry symbolic meaning in itself and override the literal meaning. Turning to the English text on the left, it seems rather unlikely that the word special is employed to create the illustrated connotations. In fact, it may even be interpreted as an indicator for the exoticness, originality and authenticity of the foreign products advertised. It has become evident that the different languages and script systems used fulfil different functions for distinct audiences. For a speakers of Urdu, the English language may be symbolic of modernity, while the presence of Urdu creates a sense of familiarity, solidarity and inclusiveness. For an audience that understands both Urdu and English, all linguistic items will have referential; yet, the meanings and interpretation of the texts may still be shaped by the mixing of languages and script.

7 Discussion Having analysed the LL of the Curry Mile both quantitatively and qualitatively, I can now provide answers to the research questions of this study. The following sections will address each question separately, which shall help to draw general conclusions with respect to language functions in the LL.

7.1 The LL and Ethnolinguistic Vitality

The first question addresses Landry and Bourhis’ understanding of the LL as a reflection of language vitality and the status of language groups inhabiting the investigated area. On the basis of the findings, it can be claimed that the investigated LL only partly reflects the assumed language situation of Rusholme. The quantitative analysis has shown that English seems to be omnipresent on official top-down signage, bottom-down signage and transgressive LL items. This mirrors the wider significance of English as a communication tool in the An- glophone environment, and it indicates that a major part of the intended audience can at least understand English. Languages other than English show an irregular presence. The analysis of multilingual signs has revealed that besides English, Arabic and Urdu are permanent constituents in diverse areas, which reflects the data from the census (NOMIS 2011). Additionally, minority languages appear on a considerable number of monolingual signs, which attaches them a relatively important and autonomous position. Yet, it was revealed that languages present in the LL do not necessarily function as a means of communication. This accounts for the fact that conclusions regarding ethnolinguistic vitality must be drawn with caution. As language in some cases functions symbolically rather than as a communication tool, signs cannot readily be regarded as a true indication of the respective language group’s vitality. It has become apparent in this paper that LL items must be related to language use in their direct and indirect environment if it is the aim to understand how language functions in a given context. A domain-related analysis of language use outside along with the investigation of linguistic choices inside individual establishments has provided remarkable insights into the use of languages in the Curry Mile and surrounding area.

7.2 Language Use and Functions across the LL The second research question of this paper focused on language use according to different LL actors, the distribution of languages across domains and a comparison of code choices between external and internal signs. It has already been illustrated that different motivations seem to influence different types of LL actors 43 in their code-choices. The qualitative analyses have suggested that for many bottom-up signs, language functions other than the referential function may for some readers be emphasised. Furthermore, the illustrated examples have sup- ported the assumption that the nature of the product or services advertised may significantly affect language choice. The presence of certain minority languages for instance seems to have beneficial effects in relation to particular products and services only, which was shown in the comparison of indoor and outdoor signs. Shops selling or repairing electronic devices, travel agencies as well as the financial sector tend to “hide” their South-Asian origin through an exclusive use of English on their façades. In other domains, however, these codes are used in salient positions clearly visible for passers-by. The analysis leads to the conclusion that certain competencies are assigned to particular cultures and their respective languages, which in turn substantiates the claims made by Kelly-Holmes (2005) and Haarmann (1989). For commercial reasons, the referential function of language seems to be subordinate in some cases and for particular readers. Us- ing language as a commodity rather than as a means of communication, “in- depth and familiar knowledge of the foreign language is neither displayed by the [sign-writer] nor assumed on the part of the addressee” (Kelly-Holmes 2005: 13). The initial interpretation regarding correlations between language distributions across domains and their use on outdoor, or respectively indoor signs have led to striking results in the micro-level analysis. Languages whose use is concentrated among particular domains, such as for instance the gastronomy sector, appear mainly outside (cf. Figure 12). As illustrated, a great majority of the restaurants and shisha bars use Arabic in prominent positions on their signs outside, but English seems to be the main means of communication inside. Likewise, Ital- ian and French are present only on external signage of certain businesses. With their focus on language-content relationships, the analyses have illustrated that the use of minority languages is in some cases confined to business names. On the other hand, a consistent use of multiple languages to convey various types of information suggests that potential readers have different first languages, and that sign-producers presuppose individual monolingualism or insuf- ficient knowledge in either of the languages used. Rather than being used for their connotational values, the codes serve as communication tools for different language groups who have no common language. Similar to what was claimed about top-down and bottom-up signs above, it can be hypothesised that particular functions are in certain locations likely to be more dominant than others. Language on external signage might be chosen at-

44 tract the audience rather than to communicate referential content, whereas codes inside the establishment are more to likely function as an actual means of communication. Reminiscent of phatic speech, which can be regarded as a precursor to further social interactions as it opens the channel for communication, signs on the façades of establishments might be intended primarily to create a social contact with the passers-by. Illustrating the emotive function of language, the expression of one’s own identity or attitudes through language can serve to appeal to and influence the viewer. Similarly, the aesthetic value of language may be of importance in attracting attention. While the phatic, conative, emotive or poetic functions of language might be foregrounded for signs on the façade of shops, it may be claimed that menus or product information displayed inside are primarily referential and informative. Such functional differences may greatly influence language choice for LL signs. Yet, it is not suggested that the functions fulfilled by external signs differ fun- damentally from the purposes of written text displayed inside. Indeed, outdoors signage is frequently designed to communicate referential content and inform customers about opening hours, the products range and the like. Similarly, signs inside may, of course, involve attention-getting, phatic elements; moreover, most interior signs are oriented towards the addressee and have the purpose of per- suading the reader and influencing their actions. Moreover, it must be re- emphasised that the functions do not exclude one another. Nevertheless, the data has revealed certain tendencies according to the relation between placement and functions of signs, which can be essential indicators for LL researchers in exploring the functions of language in the LL. Turning to the third question posed in this paper, the present research has revealed that distinct functions of language may be foregrounded for different readerships, which is why a distinction between informational and symbolic functions of language seems highly inappropriate.

7.3 The Functions of Languages as Gradient Features

The current study found that certain associations are likely to be evoked by particular languages, but languages cannot be assigned fixed sets of symbolic values. It has become clear that distinct readers may perceive language on signs differently. Considering that language knowledge, experience and attitudes differ between LL actors and readers, the character of interpretation and meaning- making is largely dependent on individuals. This implies that the referential func-

45 tion cannot be neglected per se. In other cases, linguistic items and the language as such might acquire symbolic value, to denotational meanings. Contrary to claims made in previous LL research (cf. Scollon/Wong Scollon 2003; Finzel 2012), the findings of this study suggest that the relation between the functions of language in the LL is not dichotomous. The informational and symbolic functions of language are not to be regarded as clearly definable categories that can be strictly separated from each other. Rather, they are closely interrelated. Despite the general absence of clear-cut boundaries between informational and symbolic functions, it has become evident that if one considers a variety of factors, one can make assumptions about the sign-producer’s motivations that leading to certain code choices. Some signs seem to be more informational, whereas other uses of language on signs foreground symbolic values. Interpret- ed from a particular reader-perspective, one of the language functions may be more dominant for a given text part on a sign. Hence, LL items can in a given context be positioned on a graduated scale. Such a scale reflects the role of languages in conveying information, and indicates the extent to which the code is considered a means of communication for the intended readerships. Not only the presence and combinations of languages on the sign, but also the informational values of languages and possibly visual features need to be considered. Fur- thermore, language use in the sign’s direct and indirect context must be taken into account. On one end of the continuum, languages on signs serve mainly to communicate content and provide readers with information. In these cases, code choice is determined by assumed linguistic competencies of the targeted readership. On the sign visible in Figure 27, different types of information are given in two languages in order to ensure that intended audiences comprehend the information to be conveyed. All relevant information is equally presented in both languages, and also further signs of the business are characterized Figure 27. Razan Travel by duplicating multilingualism.

On the other end of the scale, certain code choices seem to be driven by connotations rather than denotation. This applies for instance to signs on which only the shop name is written in a language different from the main text. In Figure 28a and Figure 28b, it becomes evident that the type and amount of information obtainable in the languages differ. It seems that French is chosen for associated values rather than for its referential function.

Figure 28a. Mille et Une Nuits 1

Figure 28b. Mille et Une Nuits 2

Similar to language use on signs of “Caffè La Fresco” illustrated above, linguistic items are intended to evoke values and express emotions or identity rather than to communicate factual information. For the signs presented above, an analysis of the language-content relationships and a consideration of language use in their direct context makes an interpretation of the LL actors’ potential motivations relatively clear. Given the subjective character of interpretation, however, it is fundamental to note that the starting point for an investigation must be the consideration of the reader-perspective and the respective language knowledge and attitudes.

Figure 29. Falafel

Figure 29 shows a sign that informs the Arabic-speaking readership about the menu of the restaurant “Falafel”. While the linguistic items fulfil a referential function for Arabic readers, it might evoke metonymic links with the respective cultures for a European audience, enhancing the product’s authenticity. Such a scale can only attempt to catch the functions of certain languages from the perspective of a particular language group. It can indicate what role the languages are possibly intended to serve, but it does not reveal which meanings are eventually ascribed by an individual reader. Written texts displayed in the LL do not fall neatly into either of the categories; depending on the individual, they may be perceived as more prototypical, i.e. closer to one pole than to the other. In the preceding chapters, it has been shown that the functions of languages in the LL cannot be clearly determined. However, it has become evident that all LL items are somehow indexical to any viewer.

8 Conclusion

Returning to the central questions posed in this paper, it is now possible to state that functions of languages visible in the LL cannot be unambiguously determined. It has been revealed that there are no clear-cut boundaries separating informational and symbolic uses of language, as the functions of language may change with each change of perspective. The objective of this study was to increase the understanding of the roles of language on signs in the public sphere. The investigation of the LL of the Curry Mile in Manchester and the surrounding area has offered profound insights into written language use in a multilingual context. 48

This paper began by recapturing relevant research in the field of LL, paying particular attention to the analysis of language functions on signs. A determina- tion of the weaknesses found in previous theories has led to the consideration of the varied functions of language, as well as the adoption of a cognitive sociolinguistic perspective. Along with the understanding of communication as a dynamic and context-dependent process of meaning-negotiation, they have provided a sound basis for the framework put forward in this paper. The participant-oriented and contextualised approach, which combines quantitative and qualitative methods, has proven particularly valuable to adequately answer the research questions posed in this paper. One of the conclusions drawn is that linguistic choices for signs do not necessarily correspond to the intended reader’s assumed or the LL writer’s own language proficiencies. Cultural stereotypes, language attitudes and the prestige or status of languages affects linguistic choices for signs displayed in the public sphere. Moreover, languages may serve to express or construct identity. Although the investigated LL seems to generally reflect the linguistic situation of the area, the research has shown that the vitality of a language cannot be as- sessed by looking at external signage in the public sphere. An important aspect to be investigated is the referential content of the texts. A detailed analysis of the language-content relationship helps to determine which languages are intended to function as primary means of communication. Furthermore, the consideration of multimodal ways of meaning-making is necessary for a deeper understanding of language choices in the LL. In addition, a comparison of linguistic choices on bottom-up outdoors signage with language use inside the establishments was particularly revealing. It has become evident that the LL items on the façades do not necessarily indicate the languages that actually function as means of communication. Another major find- ing was that striking results in this regard correlate with results from the domain- related analysis. Particular languages tend to be displayed on the façades of certain types of business, suggesting a connection between languages and the nature of products or services offered. However, it would be wrong to assume that certain languages can be linked to particular sets of symbolic values, neither are they limited to certain functions. First, it must be noted that the relation between language and ethnicity or culture remains subjective, since there is no categorical, necessary connection between the two. Second, the associations a language may evoke and the meanings a sign carries are also highly subjective and dependent on the individual’s linguistic

49 knowledge, experience and attitudes. In a process of negotiation and interpretation of meaning, one or several functions may be ascribed to language in the LL. Yet, the sign-writers’ intentions are not necessarily in accord with the interpretation eventually made by the individual reader. Investigating language-in-use, it is thus essential to analyse the text both from the producer’s, as well as the receiv- ing reader’s perspective and relate it to the sign’s context. Illustrating that the relationship between the informational and symbolic functions of language on signs cannot be conceived as dichotomous, the present study makes a major contribution to LL research. Not only does it enhance a more adequate understanding of the dynamic functions of written language in the public sphere, but it also provides a methodological framework that can be profit- ably applied in future LL studies. In a multilingual environment such as the Curry Mile, different languages become re-contextualized when they come into contact, and they may have a powerful impact even on the unconscious passers-by.

9 Bibliographical References

Agnihotri, Rama K. & Kay McCormick. 2010. “Language in the Material World: Multilinguality in Signage.” The International Multilingual Research Journal, 4. 55-81.

Backhaus, Peter. 2007. Linguistic Landscapes: A Comparative Study of Urban Multilingualism in Tokyo. Clevedon: Multilingual Matters.

Backhaus, Peter. 2006. "Multilingualism in Tokyo. A Look into the Linguistic Landscape". International Journal of Multilingualism, 3/1. 52–66.

Barni, Monica & Carla Bagna. 2009. “A Mapping Technique and the Linguistic Landscape.” In: Elana Shohamy & Durk Gorter (eds.). Linguistic Landscape. Expanding the Scenery. London/New York: Routledge: 126-140.

Ben-Rafael, Eliezer; Shohamy, Elana; Amara, M.Hasan & Nira Trumper-Hecht. 2006. „Linguistic Landscape as Symbolic Construction of the Public Space. The Case of Israel.” International Journal of Multilingualism, 31. 7-30.

Blommaert, Jan, James Collins & Stef Slembrouck. 2005. “Spaces of Multi- lingualism.” Language & Communication, 25/3. 197-216.

Bublitz, Wolfram. 2012. "From Speaker and Hearer to Chatter, Blogger and User. The Changing Metacommunicative Lexicon in Computer-Mediated Commmunication." In: Busse, U. & A. Hübler (eds.). Investigations into the Meta-Communicative Lexicon of English. A Contribution to Historical Prag- matics. Amsterdam/ Philadelphia: Benjamins, 151-176.

Bublitz, Wolfram 2009. Englische Pragmatik. Eine Einführung. Berlin: Eric Schmidt.

Bublitz, Wolfram 2008. "Sailing the Islands or Watching from the Dock. The Treacherous Simplicity of a Metaphor. How we handle 'new (electronic) Hy pertext' vs 'old (printed) Text'." In: Gerbig, A. & O. Mason (eds.) Language, People, Numbers. Corpus Linguistics and Society. Language and Comput- ers: Studies in Practical Linguistics 64. Amsterdam: Rodopi, 249-273.

Bublitz, Wolfram. 2006. “It utterly Boggles the Mind. Knowledge, Common Ground and Coherence” In: Pishwa, H. (ed.). Language and Memory. As- pects of Knowledge Representations. Berlin: de Gruyter. 359-386.

Bühler, Karl. 1965. Sprachtheorie. Die Darstellungsfunktion der Sprache. Stuttgart: Gustav Fischer.

Cenoz, Jasone & Durk Gorter. 2006. “Linguistic Landscape and Minority Languages.” International Journal of Multilingualism, 3/1. 67-80.

Cheshire, Jenny & Lisa-Marie Moser. 1994. “English as a Cultural Symbol. The Case of Advertisements in French-Speaking Switzerland.” Journal of Multi- lingual and Multicultural Development, 15/6. 451-469. URL: http://webspace.qmul.ac.uk/jlcheshire/pdf%20papers/Cheshire%20and%20 Moser.pdf (last accessed February 14, 2014).

Collins, James & Stef Slembrouck. 2007. “Reading Shop Windows in Globalized Neighborhoods. Multilingual Literacy Practices and Indexicality.” Journal of Literacy Research, 39. 335 – 355.

Coupland, Nikolas. 2012. “Bilingualism on Display. The Framing of Welsh and English in Welsh Public Spaces.” Language in Society, 41. 1–27.

Coupland, Nikolas. 2010. “Welsh Linguistic Landscapes ‘From Above’ and ‘From ‘Below’.” In: Jaworski, Adam & Crispin Thurlow (eds.). Semiotic Landscapes. Text, Image, Space. London: Continuum, 77–101.

Donakey, Andrea 2007. Language Planning and Policy in Manchester. Published MA Dissertation. University of Manchester. URL: http://www.findthatpdf.com/search-35351072-hPDF/download-documents- ad-dissertation-pdf.htm (last accessed February 14, 2014).

Finzel, Anna M. 2012. “English in the Linguistic Landscape of Hong Kong. A Case Study of Shop Signs and Linguistic Competence.” Published Magister Thesis. University of Potsdam. URL: http://opus.kobv.de/ubp/volltexte/2013/6412/pdf/finzel_diss.pdf (last accessed March 8, 2014). Gorter, Durk. 2006. “Introduction. The Study of the Linguistic Landscape as a New Approach to Multilingualism. Clevedon, Buffalo: Multilingual Approach to Multilingualism.” In: Gorter, Durk (ed.) Linguistic Landscape. A New Approach to Multilingualism. Clevedon, Buffalo: Multilingual Matters, 1-6.

Gumperz, John J. 1982. Discourse Strategies. Cambridge: Cambridge UP.

Haarmann, Harald. 1989. Symbolic Values of Foreign Language Use. From the Japanese Case to a General Sociolinguistic Perspective. New York: Mouton De Gruyter.

Hoffmann, Christian R. (ed.). 2010. Narrative revisited. Telling a Story in the Age of New Media. Amsterdam/Philadelphia: Benjamins.

Huebner, Thom. 2006. “Bangkok’s Linguistic Landscapes. Environmental Print, Codemixing and Language Change.” International Journal of Multilingualism, 3/1. 30-57.

Jakobson, Roman. 1960. „Linguistics and Poetics.” In: Thomas A. Sebeok (ed.). Style in Language. Cambridge: MIT-Press, 350-377.

Jaworski, Adam & Crispin Thurlow. 2010. “Introducing Semiotic Landscapes.” In: Semiotic Landscapes. Language, Image, Space. London: Continuum, 1–40.

Kallen, Jeffrey. 2010. “Changing Landscapes: Language, Space and Policy in the Dublin Linguistic Landscape.” In: Jaworski, Adam & Crispin Thurlow (eds.). Semiotic Landscapes. Language, Image, Space, 41-58. London: Continuum.

Kelly-Holmes, Helen. 2005. Advertising as Multilingual Communication. Basinstoke: Palgrave MacMillan.

Kövecses, Zoltán. 2010. Metaphor. A Practical Introduction. 2nd ed. Oxford: Oxford UP.

Kress, Gunther & Theo Van Leeuwen. 2006. Reading Images. The Grammar of Visual Design. London/New York: Routledge.

Kristiansen, Gitte & René Dirven. 2008. “Cognitive Sociolinguistics. Rationale, Methods and Scope.” In: Gitte Kristiansen & René Dirven (eds.). Cognitive Sociolinguistics. Language Variation, Cultural Models, Social Systems. Ber- lin/New York: Mouton de Gruyter, 45-91.

Lakoff, George & Mark Johnson. 2003. Metaphors We Live By. 2nd ed. London: University of Chicago Press.

Landry, Rodrigue & Bourhis, Richard Y. 1997. “Linguistic Landscape and Ethn linguistic Vitality. An Empirical study”. Journal of Language and Social Psy- chology, 16/1. 23-49.

Leeman, Jennifer & Gabriella Modan. 2009. “Commodified Language in China- town. A Contextualized Approach to Linguistic Landscape.” Journal of Socio- linguistics, 13/3. 332-362.

Malinowski, David. 2009. “Authorship in the Linguistic Landscape. A Multimodal Performative View.” In: Shohamy, Elana & Durk Gorter (eds.). Linguistic Landscape: Expanding the Scenery. London/New York: Routledge, 107-125.

Matras, Yaron. 2008. “Defining ‘Everyday Multilingualism’.” Everyday Multi- linguallism Conference Report. UNESCO.

NOMIS. 2011a. “The 2011 Census: Ethnic Group Rusholme.” Office for National Statistics. URL: http://www.nomisweb.co.uk/census/2011 (last accessed March 8, 2014).

NOMIS. 2011b. “The 2011 Census: Main Language Rusholme.” Office for N tional Statistics. URL: http://www.nomisweb.co.uk/census/2011 (last accessed March 8, 2014).

Pennycook, Alastair. 2009. “Linguistic Landscapes And The Transgressive Semi- otics Of Graffiti.” In: Shohamy Elana and Durk Gorter (eds). Linguistic Land- scape. Expanding the Scenery. London/New York: Routledge, 302-312.

Piller, Ingrid. 2003. “Advertising as a Site of Language Contact.” Annual Review of Applied Linguistics, 23. 170-183.

Reh, Mechthild. 2004. “Multilingual Writing. A Reader-Orientated Typology – With Examples from Lira Municipality, Uganda.” International Journal of the Soci- ology of Language, 170. 1-41.

Rosenbaum, Yehudit, Nadel, Elizabeth, Cooper, Robert L., & Fishman, Joshua A. 1977. “English on Keren Kayemet Street.” In: Joshua A. Fishman, Robert L. Cooper & Andrew W. Conrad (eds.). The Spread of English. Rowley: New- bury House, 179–196.

Scollon, Ron & Suzie Wong Scollon. 2003. Discourses in Place. Language in the Material World. London/New York, NY: Routledge.

Seargeant, Philip. 2012. “Between Script and Language: The Ambiguous Ascrition of ‘English’ in the Linguistic Landscape.” In: Christine Hélot, Monica Barni, Rudi Janssens & Carla Bagna (eds.). Linguistic Landscapes, Multilin- gualism and Social Change. Frankfurt Am Main: Peter Lang, 187–200.

Sebba, Mark. 2012. “Multilingualism in Written Discourse. An Approach to the Analysis of Multilingual Texts.” International Journal of Bilingualism, 17. 97- 118.

Shohamy, Elana & Durk Gorter. 2009. Linguistic Landscape. Expanding the Scenery. London/New York: Routledge.

Shohamy, Elana & Shoshi Waksman. 2009. “Linguistic Landscape as an Ecolo ical Arena: Modalities, Meanings, Negotiations, Education.” In: Shohamy, Elana and Durk Gorter (eds). Linguistic Landscape. Expanding the Scenery. London/New York: Routledge, 313-331.

Spolsky, Bernard & Robert L. Cooper. 1991. The Languages of Jerusalem. New York: Oxford University Press. Taylor-Leech, Kerry J. 2012. “Language Choice as an Index of Identity. Linguistic Landscape in Dili, Timor-Leste.” In- ternational Journal of Multilingualism, 9/1. 15-34.

10 Appendix

Language Use According to Domains

Sector n Languages Used (predominant language in bold type) 1. Automotive sector 3 English 2. PC & Electronics 10 English, Arabic 3. Travel & Cargo 9 English, Arabic, Kurdish, 4. Finances 7 English, Kurdish, Arabic, Urdu 5. Solicitors 5 English 6. Medical Sector 6 English, Arabic, Chinese 7. Property, Housing 5 English, Arabic, Spanish 8. Jewellery 8 English 9. Hair & Beauty 9 English, French, Arabic 10. Clothing 14 English, Kurdish, Arabic 11. Other Services 5 English, Arabic 12. Convenience Store 8 English 13. Groceries & Supermarkets 12 English, Arabic, Kurdish, Urdu (Ar- abic + Roman script) 14. Gastronomy Businesses 84 English, Arabic, Urdu, Farsi, Italian, Total24 Hindi, Chinese, Spanish, Somali, Russian, Malay, Kurdish 14.1 Cafés 6 English, Italian 14.2 Pubs 3 English 14.3 Restaurants 41 English, Arabic, Urdu, Farsi, Italian, Hindi, Chinese, Spanish, Somali, Russian, Malay, Kurdish 14.4 Fast Food & Take- 16 English, Urdu, Farsi aways 14.5 Shisha Bars 18 Arabic, English, Farsi Figure 11. Language Use According to Domains

Figure 12. Language Use Outside versus Inside (On electronic storage) Figure 30. Bottom-Up Quantitative Analysis According to Domains (on electronic storage medium

24 During the fieldwork, a number of language use patterns became evident among certain types of gastronomy businesses. Thus, it was decided to subdivide this sector into further categories. 55